Effects of Heating Conditions on the Glass Transition Parameters of

ARTICLE pubs.acs.org/JAFC

Effects of Heating Conditions on the Glass Transition Parameters of Amorphous Sucrose Produced by Melt-Quenching Joo Won Lee Department of Food Science and Human Nutrition, University of Illinois at Urbana
Champaign, 399A Bevier Hall, 905 South Goodwin Avenue, Urbana, Illinois 61801, United States

Leonard C. Thomas DSC Solutions LLC, 27 East Braeburn Drive, Smyrna, Delaware 19977, United States

Shelly J. Schmidt* Department of Food Science and Human Nutrition, 367 Bevier Hall, University of Illinois at Urbana
Champaign, 905 South Goodwin Avenue, Urbana, Illinois 61801, United States ABSTRACT: This research investigates the effects of heating conditions used to produce amorphous sucrose on its glass transition (Tg) parameters, because the loss of crystalline structure in sucrose is caused by the kinetic process of thermal decomposition. Amorphous sucrose samples were prepared by heating at three different scan rates (1, 10, and 25 °C/min) using a standard differential scanning calorimetry (SDSC) method and by holding at three different isothermal temperatures (120, 132, and 138 °C) using a quasi-isothermal modulated DSC (MDSC) method. In general, the quasi-isothermal MDSC method (lower temperatures for longer times) exhibited lower Tg values, larger ΔCp values, and broader glass transition ranges (i.e., Tg end minus Tg onset) than the SDSC method (higher temperatures for shorter times), except at a heating rate of 1 °C/min, which exhibited the lowest Tg values, the highest ΔCp, and the broadest glass transition range. This research showed that, depending on the heating conditions employed, a different amount and variety of sucrose thermal decomposition components may be formed, giving rise to wide variation in the amorphous sucrose Tg values. Thus, the variation observed in the literature Tg values for amorphous sucrose produced by thermal methods is, in part, due to differences in the heating conditions employed. KEYWORDS: glass transition (Tg), sucrose, thermal decomposition, standard DSC, quasi-isothermal MDSC

’ INTRODUCTION Amorphous sugars are essential ingredients in many food and pharmaceutical products due to their useful properties, such as encapsulation ability, high dissolution rate, and high solubility.1,2 Examples of food ingredients and products containing amorphous sugars are spray-dried flavors, boiled sweets, milk powder, coffee powder, infant formula, and fondants,3,4 and examples of pharmaceutical ingredients and products are excipients for large-scale tablet and capsule manufacturing5
11 and fillers in chewable tablets.12,13 However, amorphous sugars are known to undergo structural changes induced by heat and/or moisture uptake, leading to undesirable events, such as stickiness, caking, and collapse and recrystallization.1,10,14 These undesirable events are detrimental to product quality attributes, such as texture, taste, aroma retention, shelf life, and delivery and stability characteristics of drug products, which are critical to final product acceptance. These stability concerns associated with amorphous sugar based products are of practical and economical importance; thus, they have been the subject of much research over the past several years. A parameter of critical importance to the stability of amorphous materials is the glass transition temperature (Tg), which is the temperature at which a reversible transition occurs between the solid amorphous (glassy) state and the supercooled liquid r 2011 American Chemical Society

(rubbery) state. Because the glassy to rubbery transition takes place over a range of temperatures, rather than at a single temperature, it is important to report the onset, midpoint, and end point of the transition and to specify the method and parameters used to obtain the Tg values.15 Upon inspection of the literature-reported Tg values for the same amorphous material, differences are often observed. For example, the Tg onset for amorphous sucrose reported by Roos16 was 62 °C, whereas Slade and Levine17 reported a value of 52 °C. The dissimilarity between reported Tg values for the same sugar has been ascribed to differences in (1) residual water content, (2) sample handling techniques (e.g., preparation methods), (3) Tg measurement techniques, and (4) Tg analysis conditions and methods.16,18,19 The factor influencing Tg of interest in this study is the heating conditions used to produce the amorphous sucrose. The heating conditions used are important because the loss of crystalline structure (usually termed melting) in sucrose is due to the kinetic process of thermal decomposition (termed apparent melting), rather than thermodynamic melting, as shown by Lee et al.20,21 Received: December 16, 2010 Revised: March 2, 2011 Accepted: March 7, 2011 Published: March 07, 2011 3311

dx.doi.org/10.1021/jf104853s | J. Agric. Food Chem. 2011, 59, 3311–3319

Journal of Agricultural and Food Chemistry Sucrose melting is a time
temperature combination process; that is, in general, thermal decomposition is more extensive under lower temperature
longer time conditions than under higher temperature
shorter time conditions. Thus, depending on the heating condition employed, different amounts and types of thermal decomposition components are formed in the amorphous sucrose and, consequently, give rise to different Tg values. It is important to note that the amorphous sucrose produced by melt-quenching, under all heating conditions (even at temperatures below the literature-reported melting temperature for sucrose), is chemically different from the starting crystalline sucrose, due to the formation of thermal decomposition components produced during heating. The amorphous sample is not just amorphous sucrose, but rather amorphous sucrose plus the resultant decomposition components. However, for convenience and consistency with the literature, the amorphous product produced by melt-quenching sucrose will be referred to as “amorphous sucrose” throughout this paper. Both Vanhal and Blond,3 studying sucrose, and Jiang et al.,22 studying sucrose, glucose, and fructose, applied different heating conditions to these sugars to explore the relationship between thermal decomposition and Tg values. Vanhal and Blond3 investigated the Tg values for amorphous sucrose prepared by various heating conditions (i.e., final heating temperature, the residence time at the final temperature, and heating rate). The Tg values for amorphous sucrose decreased in the final heating temperature range of 190
210 °C and then increased in the final heating temperature range of 215
225 °C. When sucrose was held for various lengths of time at its melting peak temperature, 190 °C, the Tg values initially decreased and then increased with increasing residence time. Vanhal and Blond3 attributed the trend of decreasing and then increasing Tg values for amorphous sucrose in these experiments to the formation of different types and amounts of thermal decomposition components. That is, the decrease in Tg values was due to the formation of small molecular weight components via bond breaking (under relatively mild heating conditions), and the increase in Tg values was due to the formation of various high molecular weight components via polymerization (under relatively severe heating conditions). Jiang et al.22 reported a similar trend of heating condition dependent changes in Tg values for fructose and glucose as well as sucrose. Additionally, Jiang et al.22 suggested that water formed via dehydration, which occurs during sugar decomposition, as another possible explanation for decreasing Tg values. The heating conditions (i.e., holding sugar samples for different residence times at different temperatures) employed by Vanhal and Blond3 and Jiang et al.22 were carried out at and above the melting peak temperature of the sugars. However, because the loss of crystalline structure in sucrose is due to the kinetic process of thermal decomposition,20,21 what still needs to be assessed is the effect on Tg values of amorphous sucrose produced under low temperature
long time conditions. Vanhal and Blond3 also investigated the effect of heating rates on the Tg values of amorphous sucrose. These researchers found that Tg mid increased from approximately 61 to 71 °C as a function of heating rate, from 5 to 40 °C/min, respectively. These researchers concluded that at slower heating rates the Tg value was lower because at slower heating rates the sucrose remains for a longer time at each temperature, which resulted in the formation of more thermal decomposition components and, consequently, lower Tg values. However, the effect of the heating rate on the amorphous sucrose Tg values could be confounded

ARTICLE

with their selection of a constant final heating temperature of 200 °C, which was applied to all heating rates (5, 10, 20, 30, and 40 °C/min). In DSC analysis, the temperature at which the endothermic melting peak for sucrose, glucose, and fructose is completed depends on the heating rate. In other words, as the heating rate increases, the temperature at which the endothermic melting curve is completed increases. As discussed previously, this is because the loss of crystalline structure of these sugars is due to the kinetic process of thermal decomposition (termed apparent melting), rather than thermodynamic melting. For example, as shown in Figure 1, the final heating temperature required for obtaining the entire apparent melting curve for sucrose was approximately 184 °C at a heating rate of 1 °C/min, whereas it was approximately 206 °C at a heating rate of 25 °C/min. Thus, if a constant final heating temperature (e.g., 206 °C) is applied to both heating rates, the amorphous sucrose heated at a rate of 1 °C/min would still be exposed to heat until the final heating temperature of 206 °C was reached. It remains ambiguous whether a change in the Tg values for the amorphous sucrose was caused only by the heating rate effect or also by the additional heat received to reach the final heating temperature. Therefore, the objective of this research was to determine (1) the effect of heating rate only on the Tg values for amorphous sucrose (final heating temperature determined by heating rate) using SDSC and (2) the effect on Tg values of producing amorphous sucrose under low temperature (well below the reported melting temperature for sucrose)
long time conditions using quasi-isothermal MDSC.

’ MATERIALS AND METHODS Materials. High-purity crystalline sucrose (g99.5) was purchased from Sigma-Aldrich Co. (St. Louis, MO) and used without further purification. The water content of sucrose measured by coulometric Karl Fischer titration (with Hydranal Coulomat AG solvent) was 0.004% wet basis (wb). Methods. Amorphous sucrose samples were prepared by heating at different rates (1, 10, and 25 °C/min) using SDSC and by holding at different isothermal temperatures (120, 132, and 138 °C) using quasiisothermal MDSC. A DSC Q2000 (TA Instruments, New Castle, DE), equipped with a refrigerated cooling system (RCS 90), was utilized for both SDSC and quasi-isothermal MDSC. Prior to the preparation and measurement of amorphous sucrose, the calibration for enthalpy (cell constant) and temperature was performed with indium (Tm onset of 156.60 °C, ΔH of 28.71 J/g, TA Instruments). In advance of using the quasi-isothermal MDSC, DSC and MDSC heat capacity calibrations (a modulation amplitude of (1 °C and a period of 100 s) were individually conducted using a 22.93 mg sapphire disk (TA Instrument) hermetically sealed in a pan. Hermetic aluminum Tzero pans and lids (TA Instruments) were used for all calibrations, sample preparations, and sample measurements. An empty pan was used as the reference. Dry nitrogen, at a flow rate of 50 mL/min, was used as the purge gas. Sample Preparation Using Different Heating Rates by SDSC. As previously mentioned in the Introduction, the desired final heating temperature required for obtaining an entire apparent melting peak for sucrose depends on the heating rate employed. The final heating temperature for each heating rate was determined as follows. First, the apparent melting end temperature was determined by linear extrapolation for triplicate measurements at each heating rate. Then, the final heating temperature for each heating rate was calculated as the average end temperature plus 2 standard deviations. The 2 standard deviations were added to account for run-to-run variation, ensuring complete loss of crystalline structure for each measurement. The final heating 3312

dx.doi.org/10.1021/jf104853s |J. Agric. Food Chem. 2011, 59, 3311–3319

Journal of Agricultural and Food Chemistry

ARTICLE

Figure 1. DSC thermograms for sucrose heated at three heating rates (HR), 1, 10, and 25 °C/min, to three different end temperatures using SDSC. The temperature at which the apparent melting for sucrose is completed depends on the HR employed in DSC analysis. Dotted vertical lines are the final temperatures (the average end temperature plus 2 standard deviations) employed at each HR. Arrows illustrate the additional heating that the sucrose samples would experience if the final heating temperature of 206 °C were used for all HRs. temperatures for each heating rate were 184 °C at 1 °C/min, 196 °C at 10 °C/min, and 206 °C at 25 °C/min. Crystalline sucrose (approximately 2.75 mg) hermetically sealed in a pan was heated from 25 °C to the final heating temperature calculated for each heating rate. The total heating times at each heating rate were 159 min at 1 °C/min, 78.4 min 10 °C/min, and 7.24 min at 25 °C/min. Immediately after the DSC heat flow signal reached the final heating temperature at each heating rate, the melted sucrose was cooled at a rate of 50 °C/min to
50 °C, equilibrated at
50 °C, and then reheated at a rate of 10 °C/min to 220 °C. The second DSC scan was used to measure the sample Tg parameters. No measurable endothermic melting peak was observed in any of the SDSC amorphous sucrose samples, indicating complete removal of crystalline structure. All amorphous sucrose samples prepared at each heating rate were measured in triplicate. Universal Analysis (UA) software (TA Instruments, version 4.4A) was used to determine the Tg parameters (Tg onset, Tg mid, Tg end, ΔCp) for the amorphous sucrose samples as follows: Tg onset, the temperature at the intersection of the regression line from the starting point of the specified limit (the first tangent) and the inflection tangent of the step change (the second tangent); Tg mid (by inflection), the temperature at the steepest slope on the DSC curve between the first tangent and the regression line after the transition (the third tangent); Tg end, the temperature at the intersection of the inflection tangent and the third tangent; and ΔCp, the difference between the linear extrapolated glass and the liquid heat capacity curves of material at Tg inflection.23 Sample Preparation Using Different Isothermal Temperatures by Quasi-isothermal MDSC. To prepare amorphous sucrose samples using quasi-isothermal MDSC, three isothermal temperatures (120, 132, and 138 °C) were chosen on the basis of the results of the stepwise quasiisothermal MDSC experiment for sucrose conducted in Lee et al.20,21 In the stepwise quasi-isothermal MDSC experiment, the heat capacity (Rev Cp) signal began to gradually increase at 120 °C, indicating the onset of the loss of crystalline structure in sucrose, and then slightly leveled off at

approximately 143 °C, indicating the removal of all crystalline structure in sucrose (Figure 2). Hence, 120 °C, the onset temperature of the loss of crystalline structure in sucrose, and 132 and 138 °C, the temperatures at which the loss of crystalline structure occurs relatively quickly, were selected as the isothermal temperatures. Because the loss of crystalline structure in sucrose is caused by the kinetic process of thermal decomposition, as shown in Lee et al.,20,21 the holding time required for removal of all crystalline structure in sucrose depends on the isothermal temperature employed. Thus, to determine the final holding time for each isothermal temperature, crystalline sucrose (approximately 2.75 mg) hermetically sealed in a pan was isothermally held at 120, 132, and 138 °C using quasi-isothermal MDSC (a modulation amplitude of (1 °C and a period of 100 s) well over the time at which the Rev Cp signal slightly leveled off. The point at which the Rev Cp signal slightly leveled off was regarded as the time required for removal of all crystalline structure in sucrose at each isothermal temperature. The end holding time was obtained by linear extrapolation for triplicate measurements at each isothermal temperature. Then, the final holding time for each isothermal temperature was calculated as the average end holding time plus 2 standard deviations. As in the case of SDSC, the 2 standard deviations were added to account for runto-run variation, ensuring complete loss of crystalline structure for each measurement. The final holding times for each isothermal temperature were 3014 min at 120 °C, 883 min at 132 °C, and 510 min at 138 °C (Figure 3). Crystalline sucrose (approximately 2.75 mg) hermetically sealed in a pan was isothermally held at each isothermal temperature for the corresponding time using quasi-isothermal MDSC (a modulation amplitude of (1 °C and a period of 100 s). After that, the melted sucrose was immediately cooled at a rate of 50 °C/min to
50 °C, equilibrated at
50 °C, and then reheated at a rate of 10 °C/min to 220 °C. In the second DSC scan, no measurable endothermic melting peak was observed in any of the MDSC amorphous sucrose samples. 3313

dx.doi.org/10.1021/jf104853s |J. Agric. Food Chem. 2011, 59, 3311–3319

Journal of Agricultural and Food Chemistry

ARTICLE

Figure 2. Stepwise quasi-isothermal MDSC thermogram for sucrose (a modulation amplitude of (1.0 °C, a period of 100 s, a stepwise temperature increment of 1 °C, and an isothermal time of 25 min with a data off period of 5 min for initial equilibration) used to determine the three isothermal temperatures (120, 132, and 138 °C) for preparing amorphous sucrose samples using quasi-isothermal MDSC.

Figure 3. Variations in the Rev Cp change during the loss of crystalline structure in sucrose by holding at isothermal temperatures (IT) of 120 °C for 3014 min, 132 °C for 883 min, and 138 °C for 510 min using quasi-isothermal MDSC. All amorphous sucrose samples prepared at each isothermal temperature were measured in triplicate. The Tg parameters (Tg onset, Tg mid, Tg end, ΔCp)

for the amorphous sucrose samples were analyzed using UA software as described previously. 3314

dx.doi.org/10.1021/jf104853s |J. Agric. Food Chem. 2011, 59, 3311–3319

Journal of Agricultural and Food Chemistry

ARTICLE

Table 1. Tg Parameters (Tg onset, Tg mid, Tg end, and ΔCp) for Amorphous Sucrose Samples Produced by Employing Three Heating Rates, 1, 10, and 25 °C/min, Using SDSC and by Holding at Three Different Isothermal Temperatures, 120, 132, and 138 °C, Using Quasi-isothermal MDSC, Respectively (N = 3 and n = 3, where N = Replications and n = Analysis within a Replication) Tg parametersa Tg mid (°C)

Tg end (°C)

ΔCp (J/(g 3 °C))

Tg end
Tg onset (°C)

heating condition for preparing amorphous sucrose samples

Tg onset (°C)

heating at 1 °C/min to 184 °C

25.67 ( 2.08 b4

37.45 ( 0.51 c4

44.84 ( 0.68 c5

0.8684 ( 0.0227 a1

19.17 ( 2.61 a1

heating at 10 °C/min to 196 °C

64.05 ( 1.38 a1

69.83 ( 0.51 a1

75.46 ( 1.72 a1

0.7230 ( 0.0207 b34

11.41 ( 3.01 b34

heating at 25 °C/min to 206 °C

63.83 ( 0.65 a1

68.35 ( 0.43 b1

72.94 ( 0.81 b2

0.7003 ( 0.0223 b4

9.11 ( 1.05 b4

SDSC Method

Quasi-isothermal MDSC holding at 120 °C for 3014 min holding at 132 °C for 883 min

30.92 ( 1.88 b3 35.26 ( 3.00 a2

39.59 ( 1.57 b3 44.27 ( 2.86 a2

45.36 ( 1.40 c5 50.67 ( 2.57 a3

0.7474 ( 0.0164 b3 0.7445 ( 0.0088 b3

14.44 ( 1.08 ab2 15.41 ( 1.09 a2

holding at 138 °C for 510 min

34.79 ( 0.59 a2

42.21 ( 0.40 a2

48.38 ( 0.64 b4

0.7844 ( 0.0318 a2

13.59 ( 1.20 b23

a

Means with the same letter (difference among SDSC method or difference among quasi-isothermal MDSC method) are not significantly different (p = 0.05). Means with the same number (difference among both SDSC method and quasi-isothermal MDSC methods) are not significantly different (p = 0.05).

Figure 4. Change in Tg mid (determined by the inflection method) values for amorphous sucrose samples prepared by employing three heating rates (HR), 1, 10, and 25 °C/min, using SDSC and by holding at three isothermal temperatures (IT), 120, 132, and 138 °C, using quasi-isothermal MDSC.

Statistical Analysis. Statistical analyses were carried out using SAS software (SAS 9.2 TS Level, SAS Institute Inc., Cary, NC). The general linear model (GLM) procedure was utilized for the analysis of variance (ANOVA). Tukey’s studentized range test was used to determine any significant differences between means at p = 0.05.

’ RESULTS AND DISCUSSION Effect of Heating Rate on the Tg Parameters for Amorphous Sucrose. The Tg parameters and DSC thermograms for amor-

phous sucrose samples prepared by employing three heating

rates, 1, 10, and 25 °C/min, using SDSC are given in Table 1 and Figure 4, respectively. At 1 °C/min, the amorphous sucrose exhibited the lowest Tg values and the highest ΔCp. At 10 and 25 °C/min, the amorphous sucrose showed much higher Tg values and lower ΔCp values, compared to those at 1 °C/min. However, unexpectedly, the Tg parameters for the amorphous sucrose were slightly higher at 10 °C/min than at 25 °C/min; however, their Tg parameters were very close to each other, compared to those at 1 °C/min. It was expected that the slow heating rate would have lower Tg values, because, as discussed in detail below, a slower heating rate is thought to result in more 3315

dx.doi.org/10.1021/jf104853s |J. Agric. Food Chem. 2011, 59, 3311–3319

Journal of Agricultural and Food Chemistry

ARTICLE

Table 2. Images of “As Is” and Thermally Treated Sucrose Samples

thermal decomposition than a faster heating rate. But, perhaps, the degree of thermal decomposition that occurred at the 10 °C/min heating rate was not different enough compared to the 25 °C/min heating rate to produce a distinguishable difference in the resultant Tg values. The Tg mid value (69.83 °C) for amorphous sucrose at 10 °C/min obtained in the current study was similar to literature reported Tg mid values (67.5 °C3 and 67 °C 24) for amorphous sucrose prepared and analyzed using similar conditions, except for the use of a higher final heating temperature (200 °C for both studies). In addition, as shown in Table 1, the Tg range increased with decreasing heating rates. According to Vanhal and Blond,3 as the degree of sucrose thermal decomposition increased, the Tg range (i.e., Tg end
Tg onset) became broader, indicating an increase in the complexity of the thermal decomposition components. As shown in Lee et al.,20,21 the loss of crystalline structure in sucrose is caused by the kinetic process of thermal decomposition. Hence, a slower heating rate allows sucrose to remain for a longer time at each temperature, resulting in the formation of more and different thermal decomposition components. In the early stages of sucrose thermal decomposition, small molecular weight decomposition components (e.g., glucose, fructose, 5-HMF, water, and acids) are produced through a variety of reaction pathways. As heating conditions become more severe, high molecular weight decomposition components are formed through polymerization.25
36 Consequently, in the present study, a decrease in Tg values and an increase in ΔCp as heating rate decreases are accounted for by the plasticizing effect of the small molecular weight decomposition components. The plasticizing effect of small molecular weight components, in particular, water, on Tg is a well-known property.37
41 Both Fang and Angell,42 studying fructose and galactose, and Okuno et al.,43

studying sucrose, reported that the depression of Tg values for these amorphous sugars prepared by melt-quenching may be due to the small molecular weight fragmentation products formed by the decomposition process in the vicinity of their melting temperatures. As shown in Table 1, the ΔCp at 1 °C/min is higher than the ΔCp at 10 and 25 °C/min. The higher ΔCp at 1 °C/min indicates an increase in molecular mobility, which, we hypothesize, is due to the formation of a larger amount of small molecular weight decomposition components. The presence of smaller molecules increases the space between the original molecules by blocking their attractive forces, giving rise to greater free volume and molecular mobility.18,19 In addition, it is important to note the use of different final heating temperatures for each heating rate. Different final heating temperatures at each heating rate were used to avoid the effect of the additional heat received to reach a common final heating temperature on the Tg parameters for amorphous sucrose (Figure 1). However, the amorphous sucrose prepared by heating at 1 °C/min to 184 °C exhibited the lowest Tg values and the highest ΔCp. Thus, it seems that more extensive thermal decomposition occurs at slower heating rates, at which the sucrose sample remained for a longer time at each temperature, compared to faster heating rates to higher final heating temperatures. Effect of Isothermal Temperature on the Tg Parameters for Amorphous Sucrose. The Tg parameters and DSC thermograms for amorphous sucrose prepared by holding at three isothermal temperatures, 120, 132, and 138 °C, using quasiisothermal MDSC are given in Table 1 and Figure 4, respectively. As can be seen in Table 1, the Tg values for amorphous sucrose were lower at 120 °C than at 132 and 138 °C. However, the Tg values were slightly higher at 132 °C than at 138 °C, although only statistically different for Tg end. In general, it was expected 3316

dx.doi.org/10.1021/jf104853s |J. Agric. Food Chem. 2011, 59, 3311–3319

Journal of Agricultural and Food Chemistry that lower temperature
longer time heat treatments would lead to lower Tg values (that is, until extensive polymerization would occur and then the Tg values are postulated to begin to increase), because the thermal decomposition process could take place over a longer period of time. This was the case for the 120 °C isothermal temperature, but not for 132 and 138 °C. However, because both the SDSC and quasi-isothermal MDSC methods used in the present study were the minimum heating conditions required for the complete removal of sucrose crystalline structure, it seems that both methods in terms of the degree of thermal decomposition were not severe enough to cause extensive polymerization (i.e., production of a significant amount of higher molecular weight compounds), where an increase in Tg would be expected, as postulated by both Vanhal and Blond3 and Jiang and others.22 In regard to ΔCp, the amorphous sucrose at 138 °C had a slightly higher value than those at 120 and 132 °C, but the ΔCp at 132 °C was not statistically different from the value at 120 °C. Of importance to note is the variability in the data both within each isothermal temperature and between isothermal temperatures. The variation within each isothermal temperature is illustrated by the variation in the triplicate MDSC Rev Cp signals obtained at each isothermal temperature (Figure 3) and the relatively large standard deviations associated with each isothermal temperature (Table 1). The variation between isothermal temperatures is illustrated, as discussed above, by the lack of the expected trend in the Tg values for 132 and 138 °C holding temperatures. Both sources of variability are attributed to the complex, nonuniform (i.e., nonreproducible) nature of the thermal decomposition reaction. There are a number of factors that can affect both the thermal decomposition reaction pathway and rate, such as aqueous concentration (i.e., water), time, temperature, and presence and concentration of catalysts, natural salts (e.g., chlorides, nitrates), alkaline or acidic substances, reducing sugars (e.g., glucose, fructose), and buffering compounds. 27,28,30,34,44 For example, a number of publications have reported the accelerative effect of water on sucrose hydrolysis. 27,28,30 Kelly and Brown28 specifically mentioned that the presence of moisture on the surface of sugar crystals seemed to be necessary for thermal decomposition, because the hydrogen ion (Hþ) from water is required for sucrose hydrolysis. Thus, because thermal decomposition progresses through a myriad of complex reaction pathways involving a variety of chemical transformations, even a small change in the above factors could lead to different decomposition results and, in turn, increase the variability in the Tg parameters. Comparison of Tg Parameters Obtained by Heating Rate versus Isothermal Temperature. Overall, compared to the SDSC method, the quasi-isothermal MDSC method resulted in greater sucrose thermal decomposition,20,21 producing a greater amount of more diverse decomposition components as indicated by the dark brown color of the MDSC samples compared to the SDCS samples (Table 2). Thus, as expected, the quasi-isothermal MDSC amorphous sucrose showed much lower Tg values and higher ΔCp values than the SDSC amorphous sucrose, except for the amorphous sucrose at a heating rate of 1 °C/min, which took a total of only 159 min but resulted in the lowest Tg values, the highest ΔCp value, and the broadest glass transition range (Tg end
Tg onset). Perhaps the relatively slow 1 °C/min heating rate provided a similar initial thermal decomposition scenario as in the MDSC isothermal temperature experiments (low-temperature
long-time conditions), but then

ARTICLE

the higher temperatures (due to scanning up to 184 °C) served to accelerate the further stages of the thermal decomposition process. Additional research is needed to fully elucidate the observed dramatic decrease in Tg values for 1 °C/min compared to 10 and 25 °C/min. It is important to mention that an attempt was made to compare the Tg parameters obtained herein to the Tg parameters for an amorphous sucrose sample prepared by the nonthermal process of freeze-drying and analyzed using a SDSC heating rate of 10 °C/min. It was hypothesized that the freeze-dried sample would exhibit the highest Tg values, because it would have experienced no thermal decomposition, thus no lowering of the Tg values by the formation of small molecular weight components. However, unexpectedly, the freeze-dried sample Tg values (Tg onset = 53.52 °C, Tg mid = 60.68 °C, Tg end = 62.37 °C) were lower than the SDSC samples at heating rates of 10 and 25 °C/min and higher than all of the quasi-isothermal MDSC samples and the SDSC sample at 1 °C/min. The freezedried sample ΔCp value (0.8984 J/g 3 °C) was slightly higher than the SDSC sample at 1 °C/min, which was the sample with the highest thermal decomposition. This result is ascribed to the relatively high moisture content of the freeze-dried sample. Even after 14 days in a desiccator over P2O5, the moisture content of the freeze-dried sample was 2.3% (wb), which lowered the Tg values and raised the ΔCp value compared to the SDSC samples at heating rates of 10 and 25 °C/min. Further investigation is needed to obtain Tg parameters for an amorphous sucrose sample with a much lower moisture content (as close to zero as possible), which might be produced by increasing the freezedrying time or by producing a freeze-dried sample directly in a DSC cell. Conclusions. This research was performed to investigate the effect of heating conditions on the Tg parameters of amorphous sucrose prepared by melt-quenching. As mentioned in the Introduction, the amorphous sucrose produced by melt-quenching, under all heating conditions (even at temperatures below the literature-reported melting temperature for sucrose), is chemically different from the starting crystalline sucrose, due to the formation of thermal decomposition components produced during heating. The amorphous sample is not just amorphous sucrose, but rather amorphous sucrose plus the resultant decomposition components. However, for convenience and consistency with the literature, the amorphous product produced by melt-quenching sucrose was referred to as “amorphous sucrose” throughout this paper. Overall, it was found that the quasiisothermal MDSC method (lower temperatures for longer times) exhibited lower Tg values, larger ΔCp values, and broader glass transition ranges (i.e., Tg end
Tg onset) than the SDSC method (higher temperatures for shorter times), except at a heating rate of 1 °C/min, which exhibited the lowest Tg values, the highest ΔCp, and the broadest glass transition range for all heating conditions examined herein. Because the kinetic process of thermal decomposition is responsible for the loss of crystalline structure in sucrose, the observed decrease in Tg values was ascribed to the plasticizing effect of small molecular weight decomposition components. It was hypothesized that, perhaps, the relatively slow 1 °C/min heating rate provided a similar initial thermal decomposition scenario as in the MDSC isothermal temperature experiments (low temperature
long time conditions), but then the higher temperatures (due to scanning up to 184 °C) served to accelerate the further stages of the thermal decomposition process. Additional research is needed to fully 3317

dx.doi.org/10.1021/jf104853s |J. Agric. Food Chem. 2011, 59, 3311–3319

Journal of Agricultural and Food Chemistry elucidate the observed dramatic decrease in Tg values for a heating rate of 1 °C/min compared to heating rates of 10 and 25 °C/min. In addition, this research showed that the heating conditions employed to produce the amorphous sucrose are an additional important contributing factor to explain the wide variation of Tg values observed in the literature. Because glucose and fructose have also been shown to lose their crystalline structure via thermal decomposition,20,21 the effect of heating conditions on their Tg parameters is hypothesized to be similar to that on sucrose studied herein. This research is useful for better understanding the quality and stability issues associated with heat processed sugar-containing food and pharmaceutical products.

’ AUTHOR INFORMATION Corresponding Author

*Phone: (217) 333-6963. Fax: (217) 265-0925. E-mail: sjs@ illinois.edu. Funding Sources

The generous support of the Cargill Women in Science Advanced Degree Scholarship is gratefully acknowledged.

’ ACKNOWLEDGMENT We acknowledge with thanks the technical support of TA Instruments in New Castle, DE, and the analytical support of Tate & Lyle in Decatur, IL. ’ REFERENCES (1) Mathlouthi, M. Amorphous sugar. In Sucrose Properties and Applications, 1st ed.; Mathlouthi, M., Reiser, P., Eds.; Blackie Academic and Professional: Glasgow, U.K., 1995; pp 75
100. (2) Yu, L. Amorphous pharmaceutical solids: preparation, characterization and stabilization. Adv. Drug Delivery Rev. 2001, 48, 27–42. (3) Vanhal, I.; Blond, G. Impact of melting conditions of sucrose on its glass transition temperature. J. Agric. Food Chem. 1999, 47, 4285–4290. (4) van Scoik, K. G. Nucleation and crystallization phenomena in amorphous sucrose systems. Ph.D. Dissertation, University of Wisconsin, Madison, WI, 1987. (5) Ritchie, K.; Winfield, A. J. An investigation of phenobarbitone
sucrose glass systems. Int. J. Pharm. 1984, 18, 89–97. (6) Ishikawa, T.; Mukai, B.; Shiraishi, S.; Utoguchi, N.; Fujii, M.; Matsumoto, M.; Watanabe, Y. Preparation of rapid disintegrating tablet using new types of microcrystalline cellulose (PH-M series) and low substituted-hydroxypropylcellulose or spherical sugar granules by direct compression method. Chem. Pharm. Bull. 2001, 49, 134–139. (7) Shimizu, T.; Sugaya, M.; Nakano, Y.; Izutsu, D.; Mizukami, Y.; Okochi, K.; Tabata, T.; Hamaguchi, N.; Igari, Y. Formulation study for lansoprazole fast-disintegration tablet. III. Design of rapidly disintegrating tablets. Chem. Pharm. Bull. 2003, 51, 1121–1127. (8) Joshi, A. A.; Duriez, X. Added functionality excipients: an answer to challenging formulations. Pharm. Technol. 2004, Excipients, 12
19. (9) Surana, R.; Pyne, A.; Suryanarayanan, R. Effect of preparation methods on physical properties of amorphous trehalose. Pharm. Res. 2004, 21, 1167–1176. (10) Airaksinen, S.; Karjalainen, M.; Shevchenko, A.; Westermarck, S.; Lepp€anen, E.; Rantanen, J.; Yliruusi, J. Role of water in the physical stability of solid dosage formulations. J. Pharm. Sci. 2005, 94, 2147–2165. (11) Fu, Y. R.; Jeong, S. H.; Callihan, J.; Kim, J.; Park, K. Polymeric drug delivery II: Polymeric matrices and drug particle engineering. ACS Symp. Ser. 2006, No. 924, 340–351.

ARTICLE

(12) Mullarney, M. P.; Hancock, B. C.; Carlson, G. T.; Ladipo, D. D.; Langdon, B. A. The powder flow and compact mechanical properties of sucrose and three high-intensity sweeteners used in chewable tablets. Int. J. Pharm. 2003, 257, 227–236. (13) Gohel, M. C.; Jogani, P. D. A review of co-processed directly compressible excipients. J. Pharm. Pharm. Sci. 2005, 8, 76–93. (14) Gerggren, J.; Alderborn, G. Effect of polymer content and molecular weight on the morphology and heat- and moisture-induced transformation of spray-dried composite particles of amorphous lactose and poly(vinylpyrrolidone). Pharm. Res. 2003, 20, 1039-1046. (15) Schmidt, S. J. Water and solids mobility in foods. In Advances in Food and Nutrition Research; Taylor, S., Ed.; Academic Press: London, U.K., 2004; Vol. 48, pp 1
101. (16) Roos, Y. H. Food components and polymers. In Phase Transitions in Foods; Taylor, S. L., Ed.; Academic Press: San Diego, CA, 1995; pp 109
156. (17) Slade, L.; Levine, H. Beyond water activity: recent advances based on an alternative approach to the assessment of food quality and safety. Crit. Rev. Food Sci. Nutr. 1991, 30, 115–360. (18) Roos, Y. H. Glass transition temperature and its relevance in food processing. Annu. Rev. Food Sci. Technol. 2010, 1, 469–496. (19) Abiad, M. G.; Carvajal, M. T.; Campanella, O. H. A review on methods and theories to describe the glass transition phenomenon: applications in food and pharmaceutical products. Food Eng. Rev. 2009, 1, 105–132. (20) Lee, J. W.; Thomas, L. C.; Schmidt, S. J. Investigation of the heating rate dependency associated with the loss of crystalline structure in sucrose, glucose, and fructose using a thermal analysis approach (part I) J. Agric. Food Chem. 2011, 59, 684–701. (21) Lee, J. W.; Thomas, L. C.; Jerrell, J.; Feng, H.; Cadwallader, K. R.; Schmidt, S. J. Investigation of thermal decomposition as the kinetic process that causes the loss of crystalline structure in sucrose using a chemical analysis approach (part II). J. Agric. Food Chem. 2011, 59, 702–712. (22) Jiang, B.; Liu, Y.; Bhandari, B.; Zhou, W. Impact of caramelization on the glass transition temperature of several caramelized sugars. Part I: Chemical analyses. J. Agric. Food Chem. 2008, 56, 5138–5147. (23) Wungtanagorn, R.; Schmidt, S. J. Thermodynamic properties and kinetics of the physical aging of amorphous glucose, fructose, and their mixture. J. Therm. Anal. Calorim. 2001, 65, 9–35. (24) te Booy, M. P. W. M.; de Ruiter, R. A.; de Meere, A. L. J. Evaluation of the physical stability of freeze-dried sucrose containing formulations by differential scanning calorimetry. Pharm. Res. 1992, 9, 109–114. (25) Hirschm€uller, H. Chemical properties of sucrose. In Principles of Sugar Technology; Honig, P., Ed.; Elsevier Publishing: New York, 1953; Chapter 1, pp 1
17. (26) Gardiner, D. The pyrolysis of some hexoses and derived di-, tri-, and poly-saccharides. J. Chem. Soc. C 1996, 1473–1476. (27) Mauch, W. The chemical properties of sucrose. Sugar Technol. Rev. 1971, 1, 239–290. (28) Kelly, F. H. C.; Brown, D. W. Thermal decomposition and colour formation in aqueous sucrose solutions. Sugar Technol. Rev. 1978/1979, 6, 1–48. (29) Richards, G. N.; Shafizadeh, F. Mechanism of thermal degradation of sucrose: a preliminary study. Aust. J. Chem. 1978, 31, 1825–1832. (30) Richards, G. N. Initial steps in thermal degradation of sucrose. Int. Sugar J. 1986, 88, 145–148. (31) Lowary, T. L.; Richards, G. N. Effects of impurities on hydrolysis of sucrose in concentrated aqueous solution. Int. Sugar J. 1988, 88, 145
148. (32) Tomasik, P. Thermal decomposition of carbohydrates. Part I. The decomposition of mono-, di-, and oligo-saccharides. Adv. Carbohydr. Chem. Biochem. 1989, 47, 203–270. (33) Kroh, L. W. Caramelisation in food and beverages. Food Chem. 1994, 51, 373–379. (34) Eggleston, G.; Trask-Morrell, B. J.; Vercellotti, J. R. Use of differential scanning calorimetry and thermogravimetric analysis to 3318

dx.doi.org/10.1021/jf104853s |J. Agric. Food Chem. 2011, 59, 3311–3319

Journal of Agricultural and Food Chemistry

ARTICLE

characterize the thermal degradation of crystalline sucrose and dried sucrose
salt residues. J. Agric. Food Chem. 1996, 44, 3319–3325. (35) Simkovic, I.; Surina, I.; Vrican, M. Primary reactions of sucrose thermal degradation. J. Anal. Appl. Pyrolysis 2003, 70, 493–504. (36) Quintas, M.; Guimar~aes, C.; Baylina, J.; Brand~ao, T. R. S.; Silva, C. L. M. Multiresponse modeling of the caramelization reaction. Innovative Food Sci. Emerging Technol. 2007, 8, 306–315. (37) Fox, T. G.; Flory, P. J. Second-order transition temperatures and related properties of polystyrene. I. Influence of molecular weight. J. Appl. Phys. 1950, 21, 581–591. (38) Levine, H.; Slade, L. Water as a plasticizer: physic-chemical aspects of low-moisture polymeric systems. In Water Science Reviews 3; Franks, F., Ed.; Cambridge University Press: New York, 1988; pp 79
185. (39) Roos, Y. H.; Karel, M. Plasticizing effect of water on thermal behavior and crystallization of amorphous food models. J. Food Sci. 1991, 56, 38–43. (40) Roos, Y. H.; Karel, M. Water and molecular weight effect on glass transitions in amorphous carbohydrates and carbohydrate solutions. J. Food Sci. 1991, 56, 1676–1681. (41) Brent, J. L.; Mulvaney, S. J.; Cohen, C.; Bartsch, J. A. Thermomechanical glass transition of extruded cereal melts. J. Cereal Sci. 1997, 26, 301–312. (42) Fan, J.; Angell, C. A. Relaxational transitions and ergodicity breaking within the fluid state: the sugars fructose and galactose. Thermochim. Acta 1995, 266, 9–30. (43) Okuno, M.; Kishihara, S.; Otsuka, M.; Kawasaki, K. Glass transition temperature of amorphous solid formed form sucrose crystal with different melting point. Proc. Res. Soc. Jpn. Sugar Refin. Technol. 2003, 51, 7
12. (44) Eggleston, G.; Vercellotti, J. R.; Edye, L. A.; Clarke, M. A. Effects of salts on the initial thermal degradation of concentrated aqueous solutions of sucrose. J. Carbohydr. Chem. 1996, 15, 81–94.

3319

dx.doi.org/10.1021/jf104853s |J. Agric. Food Chem. 2011, 59, 3311–3319